Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prognostic Factors for Development of Subsequent Metastases in Localized Osteosarcoma: A Systematic Review and Identification of Literature Gaps

Prognostic Factors for Development of Subsequent Metastases in Localized Osteosarcoma: A... Hindawi Sarcoma Volume 2020, Article ID 7431549, 7 pages https://doi.org/10.1155/2020/7431549 Review Article Prognostic Factors for Development of Subsequent Metastases in Localized Osteosarcoma: A Systematic Review and Identification of Literature Gaps 1 1 1 2 Patrick Basile , Emily Greengard, Brenda Weigel, and Logan Spector Department of Pediatric Hematology/Oncology, University of Minnesota, Minneapolis, MN, USA Department of Epidemiology, University of Minnesota, Minneapolis, MN, USA Correspondence should be addressed to Patrick Basile; pbasile@umn.edu Received 12 December 2019; Revised 28 January 2020; Accepted 21 February 2020; Published 27 March 2020 Academic Editor: Dae-Geun Jeon Copyright © 2020 Patrick Basile et al. *is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Aim. To investigate prognostic factors in pediatric and young adult patients with localized osteosarcoma that could predict the development of subsequent pulmonary metastases and lead to an ability to risk-stratify therapy. We performed a systematic review of the literature published since January 1990 to establish common evidence-based prognostic factors. Methods. PubMed and Embase searches (Jan 1990–Aug 2018) were performed. Two reviewers independently selected papers for patients with localized osteosarcoma with subsequent metastatic development and then reviewed for quality of methods and prognostic factors. Results. Database searches yielded 216 unique results. After screening, 27 full-text articles were studied in depth, with 9 items fulfilling predetermined inclusion and exclusion criteria. Age, tumor location, tumor size/volume, and histologic response carried in- dependent prognostic value in the majority of the studies. Conclusions. Several prognostic factors seemed to be consistent amongst the studies, but the heterogeneity and smaller sizes of the study populations made pooling of results difficult. Standardization of larger patient populations and consistent definitions/cutoffs for prognostic factors are needed to further assess for consistent prognostic factors and potential predictive models to be developed. metastases are present at initial diagnosis or if they develop 1. Introduction at subsequent follow-up evaluations [5]. High-grade osteosarcoma accounts for approximately 5% of Regardless of the timing for the development of meta- all childhood malignancies with an incidence rate of 4.4 static disease, treatment for those with metastatic disease has cases per million in patients aged 0–24 [1]. Historically, been difficult. Although treatment regimens for synchro- patients with localized disease treated with only surgical nous lesions have been tolerable, they have not increased the resection had poor outcomes with a 20% 3-year disease-free survival or event-free survival rates (EFS) to nonmetastatic survival (DFS) [2]. Overall survival has increased since the percentages [6]. Current standard of care for metastatic 1970s from approximately 15%–70% with the addition of disease remains the same as localized disease with neo- adjuvant and neoadjuvant chemotherapy to local control adjuvant chemotherapy, resection of the primary tumor, and surgery [2]. One of the most important prognostic factors adjuvant chemotherapy. Although 10% of pulmonary me- impacting survival rates is the presence of metastasis [3]. *e tastases will resolve after chemotherapy [7], the majority will most likely destination for metastatic disease is the lungs, require surgical resection following the completion of although other locations such as bone are possible. Ap- chemotherapy. *is standard of care regimen has improved proximately 50–60% of those diagnosed with osteosarcoma survival to approximately 70% [2], but also has been shown develop metachronous metastases and 20% develop syn- to have significant late effects such as cardiac insufficiency, chronous metastases [4]. Survival is influenced similarly if hearing loss, and infertility [8]. 2 Sarcoma Records identified through database Additional records identified searching through other sources (n = 217) (n = 0) Records after duplicates removed (n = 216) Records screened Records excluded (n = 206) (n = 179) Full-text articles assessed Full-text articles excluded, for eligibility with reasons (n = 27) (n = 18) Studies included in qualitative synthesis (n = 9) Figure 1: PRISMA flow diagram for the systematic review detailing the database searches, the number of abstracts screened, and the full texts retrieved. Several different prognostic factors for overall survival diagnosis. Retrospective case control studies, randomized (OS) when standard of care chemotherapy regimens con- control trials, and retrospective cohort studies were in- taining platinum agents, doxorubicin, high-dose metho- cluded in the review. trexate, and ifosfamide, are used have been discussed in the Quality assessment was completed evaluating manu- literature [9–11]. Although risk factors for the development scripts based on the defined inclusion and exclusion criteria of subsequent metastases in patients have been evaluated, as well as clear cohorts in the studies that included patients these have not been systematically reviewed. that had local disease at diagnosis and developed subsequent metastases. *e qualitative review for risk factors was *e aim of our systematic review was to identify trends in the literature of the most commonly evaluated risk factors completed by Basile and Greengard with a third reviewer for metachronous development of metastases in those pa- assigned (Spector) to resolve discordant reviews. Studies tients with localized disease at diagnosis. Early identification fulfilling the specified criteria were then reviewed for as- of these patients could lead to improvement in risk strati- sociations for risk of development of subsequent pulmonary fication and judicious use of chemotherapy. metastatic lesions. 2. Methods 3. Results MEDLINE and Embase were searched for eligible studies *e search resulted in a total of 217 publications and 216 published in English between January 1990 and August after duplicates were removed. 206 records were screened 2018 and the month before the search was run. We used after having abstracts available in English and 179 were the following search strategy: Osteosarcoma AND me- excluded on abstract review. Of the 27 full text articles tastasis AND risk AND factors AND (Humans[Mesh] assessed for eligibility, 19 were excluded for not evaluating AND (infant[MeSH] OR child[MeSH] OR adolescent for subsequent metastases and not isolating analysis to [MeSH])) and limited to human studies. Reports were patients with local disease, discussing multiple tumor types included if patients were <41 years old with initial di- without isolation of osteosarcoma or not having full text agnosis of nonmetastatic high-grade osteosarcoma at available to review. *is left 9 studies included in the diagnosis. Studies were excluded if they included patients qualitative synthesis (Figure 1). Duplicate authors were ≥41 years old (regardless if age groups were reported noted, but overlapping patient populations were not present separately) or if metastatic disease was present at upon detailed review. Included Eligibility Screening Identification Sarcoma 3 Table 1: Summary of the prognostic factors evaluated in the 9 studies included in the systematic review that included univariate or multivariate analysis. Bispo Junior ´ and Hauben Ferrari et al. Zhou et al. Ferrari et al. Kong et al. [14] Kim et al. [16] Ward et al. [19] Smeland et al. [12] Camargo [15] et al. [13] [17] [18] [20] LRR MFS MFS (N � 114) MFS (N � 24) MFS (N � 365) RFI (N � 19) MFS (N � 111) EFS (N � 1395) DFS (N � 127) (N � 33) (N � 19) UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA NS NS (15; Age NS (15) — NS (15) — S (2; 15; 40) S — — — — — — — S (sex dep) S — (q5y) 21; 30) Gender NS — NS — NS — NS — — — — — — — NS NS NS — Tumor location NS — NS — S S NS — — — — — NS — S S NS — Tumor size/ S S S (s � 6 cm; S (v � 1/3 S NS S — — — — — — — — S (v) NS — volume# (v � 150 ml) (s � 15 cm) 8 cm) bone) Osteosarcoma NS — S S NS — NS — — — — — NS — S NS NS — subtype Surgical margin S S NS — — — — — — — — — S S NS NS — — Histological S NS — — S S NS — — — — — S (90; 98) S S S S S response HER-2/neu — — — — — — — — — — S — — — — — — — expression P-glycoprotein — — — — — — — — NS — — — — — — — — ErbB-2 — — — — — — — — S — — — — — — — — — p53 — — — — — — — — S — — — — — — — — — BCL-2 — — — — — — — — — — — — — — — — — — Chemotherapy S S — — — — — — — — — — — — NS S — — protocol (MTX) (MTX) Lactate — — — — — — — — — — — — — — — — S S dehydrogenase Alkaline — — — — — — — — — — — — — — — — S NS phosphatase Pathologic NS — — — NS — — — — — — — — — NS NS fracture Limb salvage vs — — NS — — — — — — — — — — — — — amputation Time to seek — — NS — — — — — — — — — — — care UVA � univariate analysis; — � not tested; MFS � metastasis-free survival; MVA � multivariate analysis; � cut off points in years in parenthesis; q5y � 5-year increments; ns � not significant (P< 0.05); EFS � event-free survival; # � measure in parenthesis (s � significant; (s) � size and (v) � volume). 4 Sarcoma Tumor size/volume showed a slightly stronger associa- 3.1. Characteristics of Included Studies and Prognostic Factors. Most studies were retrospective in nature, and sample size tion than tumor location. *ree of the studies demonstrated an increased risk of metastases after MVA. However, there varied from 19 to 2680 patients. One study [12] was a prospective study. *is led to significant heterogeneity in were differences in both the type of measurements and sample sizes and also the majority of the sample sizes were cutoffs used for the analyses. Two of the studies used volume; small. All other articles’ subjects were derived from a single however, Kong et al. [14] used a hard cutoff of 150 mL while institutional patient population or from international col- Smeland et al. [12] evaluated the relative cutoff of 1/3 the laboration datasets [12, 13]. *ere was heterogeneity volume of the involved bone. Kim et al. [16] used maximal amongst the clinical prognostic factors that were evaluated tumor diameter in their evaluation and nomogram devel- for their association with subsequent metastasis, but several opment for the risk of metachronous metastases using factors were evaluated in the majority of the studies: age, evaluation of prognostic factors at diagnosis. *e histological subtype of these high-grade lesions did gender, tumor location, tumor size/volume, histologic subtype, and histological response (Table 1). Multivariate not appear to be significant in determining the risk of subsequent metastases. Seven of the studies performed a analysis (MVA) was performed in the reviewed papers when univariate significance was noted. UVA with only two showing an impact. After MVA, only Among these common factors, there was substantial Bispo Junior ´ and Camargo [15] showed osteoblastic subtype variation in definitions and cut off values used for the to have an effect on metastasis-free survival (MFS) with measurement of variables of interest. Age ranges varied worse rates of metastatic development compared to subtypes greatly among studies. Tumor location definitions were often other than osteoblastic. *is study did have a very low very broad or very specific; for example, Kong et al. [14] used sample of 24 patients and only stratified the subtypes as osteoblastic and nonosteoblastic. humeral vs other, while Ward et al. [19] delineated areas by bone and proximal vs distal. Studies were varied on the use of either tumor size or tumor volume measurements, and the 3.4. Surgical Margin/Histological Response. An uninvolved cutoffs used to assess risk varied study to study. Histology was surgical margin on pathological review has historically been evaluated as per pathological diagnosis in all instances, but considered important to improving outcomes in osteosar- like tumor location, the types were either somewhat broad coma [13, 16, 22]. Four studies evaluated the association with (osteoblastic vs nonosteoblastic) or somewhat specific (os- Kong et al. [14] and Ward et al. [19] showing significance teoblastic vs chondroblastic vs telangiectatic etc.). Percent with both UVA and MVA. Both of these studies used a necrosis was relatively consistent using a single stratification binary approach of involved/not-involved margins, while point of 90%; however, some studies further stratified above Bispo Junior ´ and Camargo [15] evaluated margins of 2 mm 90% while others further substratified based on the Salzer- and Smeland et al. [12] used qualitative measures of “wide/ Kuntschick grading scale [21]. radical, marginal, and intralesional.” *e studies that showed significance agree with previous cooperative group 3.2. Age/Gender. Most of the studies evaluated age at di- studies that have shown that the width of the margin does agnosis as a risk factor for the development of metachronous not appear to be significant as long as the definitive biopsy/ metastases. *ese studies used a variety of age cutoffs for resection are performed at a center with orthopedic on- cology experience [23, 24]. their analysis. Kong et al. [14] and Bispo Junior ´ and Camargo [15] used a single cutoff of 15 years, while Smeland Six of the studies evaluated the prognostic factor of histological response to neoadjuvant chemotherapy, and all et al. [12] had a gender specific age cutoff of either 11 or 12 years to distinguish between children and adolescents in used a necrosis/cell death percentage of at least 90% as the cutoff for evaluation [12–14, 16, 19, 20]. Five of the studies girls and boys, respectively. Hauben et al. [13] used 5-year increments to stratify the risk of metastatic recurrence. found continued significance on MVA. Kong et al. [14] Association of age with risk of metastases development was demonstrated a change on MFS on UVA while Hauben et al. not consistent among the studies; however, it was found to [13] did not. *is prognostic factor can be difficult to be a significant factor after MVA in EFS in the one pro- evaluate due to pathologist-to-pathologist variability in as- spective study that was included [12], demonstrating an sessments. Also, given the heterogeneity of the studies and improved survival for younger patients. Sex was included in neoadjuvant therapy regimens varied (even in a single in- stitutional study such as that by Ward et al. [19]) or were not 6 of the 9 studies, but did not show any prognostic sig- nificance with metastases development in univariate analysis clarified in the article [14]. Smeland et al. [12] performed a prospective study using the current standard of care of (UVA). methotrexate, doxorubicin, and cisplatin which is accepted as the standard of care after increasing the EFS and overall 3.3. Tumor Location/Tumor Size/Histological Subtype. survival to 54% and 71%, respectively. Location of the primary tumor was also evaluated in the majority of studies (7/9), but had variance in its estimated 3.5. Other Prognostic Factors. Several other factors were impact on metachronous metastases development. Studies by Kim et al. [16] and Smeland et al. [12] showed that evaluated among the studies but were not included in more primary tumor location in the proximal humerus had an than 3 of the 9 studies. Pathologic fracture (3 studies), limb increased risk of metastases. salvage vs amputation (1 study), time to seek care (1 study), Sarcoma 5 Primary tumor burden not only had a varying definition elevated alkaline phosphatase (ALP) (1 study), and increased Bcl-2 expression and increased P-glycoprotein expression (1 in the form of measurement criteria (volume vs size) but also had varying cutoffs between studies that did use similar study) did not show any significant changes in endpoint measurements in these individual studies. forms of measurement. As mentioned earlier, some used fixed measurements (150 ml) [14] while others used relative measurements (<1/3 of involved bone) [12]. *e latter would 4. Discussion be a more appropriate measure for pediatrics, given the *e goal of this review was to systematically evaluate the differences in potential size of the patients; however, it would literature from 1990 to present to delineate the clinical/ make data gathering potentially less consistent and more pathological risk factors that could be used to potentially subjective. predict which patients with localized osteosarcoma are more Interestingly, two molecular/genetic factors were found on UVA to have significance but were only evaluated in two likely to develop subsequent metastases. *e majority of studies were retrospective in nature, while some were studies in our search. *ere are several other molecular/ genetic factors that are found in the literature to potentially designed to address prognostic factors directly and others did not have that direct intent. *ese studies had cohorts that play a role in the outcome for patients with osteosarcoma, were able to be reviewed, but not included in a meta-analysis but consensus was clearly lacking. Based on our exclusion mainly due to the heterogeneity in definitions/cutoffs of the criteria, many of these studies were omitted as there was no different prognostic factors as well as treatment regimens data/analysis present on subsequent metastatic lesion de- being different or unclear. Also, while our initial intent was velopment or patients had metastases at diagnosis. In the to concentrate on pediatric patients (i.e.,<20 years of age at studies included in our review, Ferrari et al. [17] found that p53 mutations and absent ErbB-2 were associated with diagnosis), we subsequently opted to include studies that had patient populations less than 40 years of age due to rarity of decreased recurrence free interval (RFI), but only in UVA and in a small sample size of 19 patients. Zhou et al. [18] were the disease and to help increase the yield of the literature search. *e studies were limited to primary osteosarcoma in able to demonstrate that cytoplasmic HER-2 expression on osteosarcoma increased the risk of metastasis development. order to reduce the risk of confounding by secondary dis- ease. Histological response appeared to be the most con- HER-2 is currently being evaluated as a therapeutic target for sistent factor in prognosis for MFS in our study with osteosarcoma in a phase 1 study using chimeric antigen suboptimal response to neoadjuvant therapy (<90% ne- receptor T-cells [26]. While there may be therapeutic po- crosis), carrying a significant increase in the risk of me- tential with this receptor, its inclusion as prognostic factor tastases development. *e significance of this response is for the development of metachronous metastases cannot be echoed in the recent Children’s Oncology Group Study determined based on the current literature study. Of the factors that were found to not have significance in AOST0331 as the aim of the study was to intensify therapy in those patients with suboptimal response with the addition of the literature that was reviewed, presenting with a pathologic fracture at diagnosis was somewhat of a surprise. Pathologic ifosfamide and etoposide. However, therapy intensification did not improve outcomes and led only to increased toxicity fractures have been shown to potentially increase the risk of subsequent metastases while not affecting overall survival for the subjects [12]. Other factors that demonstrated significance via MVA [27]. However, many articles were not included in our study were age, tumor location, and tumor size/volume. Age is as the metastatic status at diagnosis was not specified or difficult to determine with any consistency among the there was no distinction between metachronous and syn- studies, given the different cutoffs that were used between chronous metastases. *is significantly limits our ability to studies. Also, our inclusion of those studies with older truly assess the potential risk that pathological fracture may patients could be skewing the data in those older age cutoff have on the development of metachronous metastases. *e development of a predictive nomogram by Kim et al. groups. Tumor location was not found to be a significant factor in [16] was an encouraging exercise and took prognostic factors into consideration that were consistent in significance with the majority of the studies it was evaluated in. Kim et al. [16] and Smeland et al. [12] showed significance for tumor lo- our review. While the nomogram appeared to be accurate in predicting actual EFS, it was only bootstrapped with cohorts cation; however, Smeland et al. [12] combined both proximal humerus and proximal femur making it more difficult to from the population that was used to develop the nomo- delineate the effect. Interestingly, these were the same two gram. Although the population was from a single institution, studies that demonstrated significance with regards to age, minimizing heterogeneity in several potential confounders raising the question of whether there is an association be- also led to a smaller sample size being used as the basis for tween age of presentation and site of primary lesion. *e the nomogram. Kim et al. [28] is another Korean group that variation of definitions of location used in the studies may also attempted to use a predictive nomogram and added in the use of ALP and capsular invasion. However, they had similar have made this risk factor more difficult to evaluate. Some studies had very broad definitions with respect to location, pitfalls with a small sample size of 141 patients and again consisted of a single ethnic group [28]. External validation of while others were more specific to sections of certain bones. *ese variations may explain why poor prognostic factors, these nomograms with larger and minimally heterogeneous such as axial skeletal and pelvic primary lesions, may have populations is required. *e establishment of cooperative been underappreciated in these studies [25]. study groups such as the European and American 6 Sarcoma other, telangiectatic, small cell, and high-grade Osteosarcoma Study (EURAMOS-1) will eventually lead to more standardized study populations and allow for these surface osteosarcoma [12]. prognostic factors to be evaluated in a prospective manner, (5) Surgical margin determination should be consistent potentially serving as the basis for new nomograms or as a using the Musculoskeletal Tumor Society (MSTS) confirmatory resource for established nomograms. originally defined by Enneking et al. [30], and this *e ability to predict the risk of metastatic disease in system classifies resection margins as intralesional patients with localized osteosarcoma would be beneficial for (macroscopic or microscopic tumor at the margin), several reasons. Most notably is a nomogram could be used marginal (resection through the pseudocapsule or for treatment stratification with the potential to spare low- reactive zone around the tumor), wide (the presence risk patients the considerable toxicity associated with our of normal tissue between tumor/pseudocapsule and current standard of care regimen. However, identification of margin), or radical (entire anatomic compartment these patients has been difficult, mainly due to issues with excised). identifying consistent prognostic factors to develop a (6) Histological response cutoff for adequacy of neo- stratification nomogram similar to the one developed by adjuvant chemotherapy should continue to be 90% Kim et al. [28] and Kim et al. [16]. During our literature tumor necrosis based on pathological determination. search and review, several factors became apparent that Central review of tissue samples should be incor- emphasize potential issues when performing a systematic porated in future clinical protocols to decrease as review or potential meta-analysis of a rare disease. First, the much interobserver variability from pathologist to number of patients between the studies varies greatly and pathologist as possible. often times can be insufficient to draw significant and consistent conclusions regarding a particular prognostic *e development of metastases is a difficult obstacle to factor. *is was true for the majority of the literature in- the successful treatment of osteosarcoma not only in per- cluded in our review, except for the Smeland et al. [12] study. sistence of disease but also in the need for more intensified Large collaborative studies such as EURAMOS-1 and those therapies. Being able to predict which patients have an conducted by the Children’s Oncology Group are absolutely increased risk of subsequent metastases after upfront necessary to obtain large pools of patients with rare diseases. therapy, would potentially allow for therapy stratification While it is important to perform these prospective studies, it after local control measures. Unfortunately, the literature is is critical to perform systematic reviews/meta-analyses to lacking in large-scale studies evaluating defined prognostic maximally use data from all available studies to draw more factors, limiting the creation of a predictive model of this accurate conclusions and inform subsequent trial rare disease difficult. *e movement towards large-scale development. collaborative studies between organizations and countries *e second major issue impacting the ability to combine will hopefully allow for the development of predictive analyses of published data is the discrepancy of measure- models in the future, ultimately leading to more effective ment cutoffs and the definitions used for several of the treatment. prognostic factors. We recommend that future studies use consistent definitions of the following prognostic factors Conflicts of Interest based on our literature review and experience: *e authors declare that they have no conflicts of interest. (1) Age should be used in sex-determined breaks due to typical differences in average ages of growth accel- References eration between the sexes. Smeland et al.’s [12] age cutoffs for children, adolescents, and adults were not [1] L. Mirabello, R. J. Troisi, and S. A. Savage, “Osteosarcoma too narrow as to cause smaller sample sizes but did incidence and survival rates from 1973 to 2004,” Cancer, distinguish differences in skeletal growth/maturity vol. 115, no. 7, pp. 1531–1543, 2009. levels. Also, Jun et al. [29] did demonstrate evidence [2] M. P. Link, A. M. Goorin, A. W. Miser et al., “*e effect of of increased growth velocity at diagnosis that was adjuvant chemotherapy on relapse-free survival in patients with osteosarcoma of the extremity,” New England Journal of associated with worse outcomes. Medicine, vol. 314, no. 25, pp. 1600–1606, 1986. (2) Tumor location should have categories of femur/ [3] E. E. Pakos, A. D. Nearchou, R. J. Grimer et al., “Prognostic tibia, other limb bones to include humerus, and axial factors and outcomes for osteosarcoma: an international skeleton. Location of proximal/distal and meta- collaboration,” European Journal of Cancer, vol. 45, no. 13, physis/diaphysis should also be included. pp. 2367–2375, 2009. [4] P. A. Meyers, C. L. Schwartz, M. Krailo et al., “Osteosarcoma: (3) Tumor size should be determined by a consistent a randomized, prospective trial of the addition of ifosfamide tumor length cutoff of <8 cm or ≥8 cm [16], with a and/or muramyl tripeptide to cisplatin, doxorubicin, and relative length of the involved bone (greater than/less high-dose methotrexate,” Journal of Clinical Oncology, vol. 23, than 1/3 of the bone) also included. no. 9, pp. 2004–2011, 2005. (4) *e osteosarcoma subtype should be categorized by [5] D. D. Rasalkar, W. C. W. Chu, V. Lee, B. K. Paunipagar, the histological subtypes of conventional chondro- F. W. T. Cheng, and C. K. Li, “Pulmonary metastases in blastic, conventional osteoblastic, conventional children with osteosarcoma: characteristics and impact on Sarcoma 7 patient survival,” Pediatric Radiology, vol. 41, no. 2, [20] S. Ferrari, G. Bacci, P. Picci et al., “Long-term follow-up and pp. 227–236, 2011 Feb. post-relapse survival in patients with non-metastatic osteo- [6] M. B. Harris, P. Gieser, A. M. Goorin et al., “Treatment of sarcoma of the extremity treated with neoadjuvant chemo- therapy,” Annals of Oncology, vol. 8, no. 8, pp. 765–771, 1997. metastatic osteosarcoma at diagnosis: a pediatric oncology group study,” Journal of Clinical Oncology, vol. 16, no. 11, [21] M. Salzer-Kuntschik, G. Delling, G. Beron, and R. Sigmund, pp. 3641–3648, 1998. “Morphological grades of regression in osteosarcoma after [7] G. Bacci, M. Rocca, M. Salone et al., “High grade osteosar- polychemotherapy? Study COSS 80,” Journal of Cancer Re- coma of the extremities with lung metastases at presentation: search and Clinical Oncology, vol. 106, no. S1, pp. 21–24, 1983. treatment with neoadjuvant chemotherapy and simultaneous [22] H. S. Kim, K. H. Shin, H. Y. Kim et al., “Postoperative no- resection of primary and metastatic lesions,” Journal of mogram to predict the probability of metastasis in Enneking Surgical Oncology, vol. 98, no. 6, pp. 415–420, 2008. stage IIB extremity osteosarcoma,” BMC Cancer, vol. 14, [8] K. A. Janeway and H. E. Grier, “Sequelae of osteosarcoma p. 666, 2014. medical therapy: a review of rare acute toxicities and late [23] A. L. Jun, M. S. Kim, D. H. Kim et al., “Osteosarcoma de- effects,” 1e Lancet Oncology, vol. 11, no. 7, pp. 670–678, 2010 veloped in the period of maximal growth rate have inferior Jul. prognosis,” Journal of Pediatric Hematology/Oncology, vol. 30, [9] A. M. Davis, R. S. Bell, and P. J. Goodwin, “Prognostic factors no. 6, pp. 419–424, 2008. [24] W. F. Enneking, S. S. Spanier, and M. A. Goodman, “A system in osteosarcoma: a critical review,” Journal of Clinical On- cology, vol. 12, no. 2, pp. 423–431, 1994. for the surgical staging of musculoskeletal sarcoma,” Clinical [10] G. Bacci, A. Longhi, M. Versari, M. Mercuri, A. Briccoli, and Orthopaedics and Related Research, vol. 153, pp. 106–120, P. Picci, “Prognostic factors for osteosarcoma of the extremity 1980. treated with neoadjuvant chemotherapy,” Cancer, vol. 106, [25] S. S. Bielack, B. Wulff, G. Delling et al., “Osteosarcoma of the no. 5, pp. 1154–1161, 2006. trunk treated by multimodal therapy: experience of the Co- [11] S. S. Bielack, B. Kempf-Bielack, G. Delling et al., “Prognostic operative Osteosarcoma study group (COSS),” Medical and factors in high-grade osteosarcoma of the extremities or Pediatric Oncology, vol. 24, no. 1, pp. 6–12, 1995. trunk: an analysis of 1,702 patients treated on neoadjuvant [26] National Library of Medicine (US), “Bethesda softworks,” cooperative osteosarcoma study group protocols,” Journal of National Library of Medicine (US), Bethesda, MD, Identifier: Clinical Oncology, vol. 20, no. 3, pp. 776–790, 2002. NCT00902044, Her2 Chimeric Antigen Receptor Expressing [12] S. Smeland, S. S. Bielack, J. Whelan et al., “Survival and T Cells in Advanced Sarcoma 2009 May 14 [cited 2019 May prognosis with osteosarcoma: outcomes in more than 2000 10]; [about 4 screens]. ClinicalTrials.gov [Internet], Available from: https://clinicaltrials.gov/ct2/show/NCT00902044, 2000. patients in the EURAMOS-1 (European and American Os- teosarcoma Study) cohort,” European Journal of Cancer, [27] Y. Zhou, Q. Lu, J. Xu et al., “*e effect of pathological fractures vol. 109, pp. 36–50, 2019. on the prognosis of patients with osteosarcoma: a meta- [13] E. I. Hauben, S. Bielack, R. Grimer et al., “Clinico-histologic analysis of 14 studies,” Oncotarget, vol. 8, no. 42, parameters of osteosarcoma patients with late relapse,” Eu- pp. 73037–73049, 2017. ropean Journal of Cancer, vol. 42, no. 4, pp. 460–466, 2006. [28] S. Jasnau, U. Meyer, J. Potratz et al., “Craniofacial osteosar- [14] C.-B. Kong, W. S. Song, W. H. Cho, J. M. Oh, and D.-G. Jeon, coma,” Oral Oncology, vol. 44, no. 3, pp. 286–294, 2008. “Local recurrence has only a small effect on survival in high- [29] D. Andreou, S. S. Bielack, D. Carrle et al., “*e influence of risk extremity osteosarcoma,” Clinical Orthopaedics and Re- tumor- and treatment-related factors on the development of lated Research, vol. 470, no. 5, pp. 1482–1490, 2012. local recurrence in osteosarcoma after adequate surgery. An [15] R. Z. Bispo Junior ´ and O. P. Camargo, “Prognostic factors in analysis of 1355 patients treated on neoadjuvant Cooperative Osteosarcoma Study Group protocols,” Annals of Oncology, the survival of patients diagnosed with primary non-meta- static osteosarcoma with a poor response to neoadjuvant vol. 22, no. 5, pp. 1228–1235, 2011. chemotherapy,” Clinics (Sao Paulo), vol. 64, no. 12, [30] N. M. Marina, S. Smeland, S. S. Bielack et al., “Comparison of pp. 1177–1186, 2009. MAPIE versus MAP in patients with a poor response to [16] M. S. Kim, S.-Y. Lee, T. R. Lee et al., “Prognostic nomogram preoperative chemotherapy for newly diagnosed high-grade for predicting the 5-year probability of developing metastasis osteosarcoma (EURAMOS-1): an open-label, international, after neo-adjuvant chemotherapy and definitive surgery for randomised controlled trial,” 1e Lancet Oncology, vol. 17, AJCC stage II extremity osteosarcoma,” Annals of Oncology, no. 10, pp. 1396–1408, 2016. vol. 20, no. 5, pp. 955–960, 2009. [17] S. Ferrari, F. Bertoni, L. Zanella et al., “Evaluation of P-gly- coprotein, HER-2/ErbB-2, p53, and Bcl-2 in primary tumor and metachronous lung metastases in patients with high- grade osteosarcoma,” Cancer, vol. 100, no. 9, pp. 1936–1942, [18] H. Zhou, R. L. Randall, A. R. Brothman, T. Maxwell, C. M. Coffin, and R. E. Goldsby, “HER-2/neu expression in osteosarcoma increases risk of lung metastasis and can Be associated with gene amplification,” Journal of Pediatric Hematology/Oncology, vol. 25, no. 1, pp. 27–32, 2003. [19] W. G. Ward, K. Mikaelian, F. Dorey et al., “Pulmonary metastases of stage IIB extremity osteosarcoma and subse- quent pulmonary metastases,” Journal of Clinical Oncology, vol. 12, no. 9, pp. 1849–1858, 1994. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Sarcoma Hindawi Publishing Corporation

Prognostic Factors for Development of Subsequent Metastases in Localized Osteosarcoma: A Systematic Review and Identification of Literature Gaps

Download PDF
7 pages

Loading next page...
 
/lp/hindawi-publishing-corporation/prognostic-factors-for-development-of-subsequent-metastases-in-8HbQwq3XUD

References (31)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Patrick Basile et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1357-714X
eISSN
1369-1643
DOI
10.1155/2020/7431549
Publisher site
See Article on Publisher Site

Abstract

Hindawi Sarcoma Volume 2020, Article ID 7431549, 7 pages https://doi.org/10.1155/2020/7431549 Review Article Prognostic Factors for Development of Subsequent Metastases in Localized Osteosarcoma: A Systematic Review and Identification of Literature Gaps 1 1 1 2 Patrick Basile , Emily Greengard, Brenda Weigel, and Logan Spector Department of Pediatric Hematology/Oncology, University of Minnesota, Minneapolis, MN, USA Department of Epidemiology, University of Minnesota, Minneapolis, MN, USA Correspondence should be addressed to Patrick Basile; pbasile@umn.edu Received 12 December 2019; Revised 28 January 2020; Accepted 21 February 2020; Published 27 March 2020 Academic Editor: Dae-Geun Jeon Copyright © 2020 Patrick Basile et al. *is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Aim. To investigate prognostic factors in pediatric and young adult patients with localized osteosarcoma that could predict the development of subsequent pulmonary metastases and lead to an ability to risk-stratify therapy. We performed a systematic review of the literature published since January 1990 to establish common evidence-based prognostic factors. Methods. PubMed and Embase searches (Jan 1990–Aug 2018) were performed. Two reviewers independently selected papers for patients with localized osteosarcoma with subsequent metastatic development and then reviewed for quality of methods and prognostic factors. Results. Database searches yielded 216 unique results. After screening, 27 full-text articles were studied in depth, with 9 items fulfilling predetermined inclusion and exclusion criteria. Age, tumor location, tumor size/volume, and histologic response carried in- dependent prognostic value in the majority of the studies. Conclusions. Several prognostic factors seemed to be consistent amongst the studies, but the heterogeneity and smaller sizes of the study populations made pooling of results difficult. Standardization of larger patient populations and consistent definitions/cutoffs for prognostic factors are needed to further assess for consistent prognostic factors and potential predictive models to be developed. metastases are present at initial diagnosis or if they develop 1. Introduction at subsequent follow-up evaluations [5]. High-grade osteosarcoma accounts for approximately 5% of Regardless of the timing for the development of meta- all childhood malignancies with an incidence rate of 4.4 static disease, treatment for those with metastatic disease has cases per million in patients aged 0–24 [1]. Historically, been difficult. Although treatment regimens for synchro- patients with localized disease treated with only surgical nous lesions have been tolerable, they have not increased the resection had poor outcomes with a 20% 3-year disease-free survival or event-free survival rates (EFS) to nonmetastatic survival (DFS) [2]. Overall survival has increased since the percentages [6]. Current standard of care for metastatic 1970s from approximately 15%–70% with the addition of disease remains the same as localized disease with neo- adjuvant and neoadjuvant chemotherapy to local control adjuvant chemotherapy, resection of the primary tumor, and surgery [2]. One of the most important prognostic factors adjuvant chemotherapy. Although 10% of pulmonary me- impacting survival rates is the presence of metastasis [3]. *e tastases will resolve after chemotherapy [7], the majority will most likely destination for metastatic disease is the lungs, require surgical resection following the completion of although other locations such as bone are possible. Ap- chemotherapy. *is standard of care regimen has improved proximately 50–60% of those diagnosed with osteosarcoma survival to approximately 70% [2], but also has been shown develop metachronous metastases and 20% develop syn- to have significant late effects such as cardiac insufficiency, chronous metastases [4]. Survival is influenced similarly if hearing loss, and infertility [8]. 2 Sarcoma Records identified through database Additional records identified searching through other sources (n = 217) (n = 0) Records after duplicates removed (n = 216) Records screened Records excluded (n = 206) (n = 179) Full-text articles assessed Full-text articles excluded, for eligibility with reasons (n = 27) (n = 18) Studies included in qualitative synthesis (n = 9) Figure 1: PRISMA flow diagram for the systematic review detailing the database searches, the number of abstracts screened, and the full texts retrieved. Several different prognostic factors for overall survival diagnosis. Retrospective case control studies, randomized (OS) when standard of care chemotherapy regimens con- control trials, and retrospective cohort studies were in- taining platinum agents, doxorubicin, high-dose metho- cluded in the review. trexate, and ifosfamide, are used have been discussed in the Quality assessment was completed evaluating manu- literature [9–11]. Although risk factors for the development scripts based on the defined inclusion and exclusion criteria of subsequent metastases in patients have been evaluated, as well as clear cohorts in the studies that included patients these have not been systematically reviewed. that had local disease at diagnosis and developed subsequent metastases. *e qualitative review for risk factors was *e aim of our systematic review was to identify trends in the literature of the most commonly evaluated risk factors completed by Basile and Greengard with a third reviewer for metachronous development of metastases in those pa- assigned (Spector) to resolve discordant reviews. Studies tients with localized disease at diagnosis. Early identification fulfilling the specified criteria were then reviewed for as- of these patients could lead to improvement in risk strati- sociations for risk of development of subsequent pulmonary fication and judicious use of chemotherapy. metastatic lesions. 2. Methods 3. Results MEDLINE and Embase were searched for eligible studies *e search resulted in a total of 217 publications and 216 published in English between January 1990 and August after duplicates were removed. 206 records were screened 2018 and the month before the search was run. We used after having abstracts available in English and 179 were the following search strategy: Osteosarcoma AND me- excluded on abstract review. Of the 27 full text articles tastasis AND risk AND factors AND (Humans[Mesh] assessed for eligibility, 19 were excluded for not evaluating AND (infant[MeSH] OR child[MeSH] OR adolescent for subsequent metastases and not isolating analysis to [MeSH])) and limited to human studies. Reports were patients with local disease, discussing multiple tumor types included if patients were <41 years old with initial di- without isolation of osteosarcoma or not having full text agnosis of nonmetastatic high-grade osteosarcoma at available to review. *is left 9 studies included in the diagnosis. Studies were excluded if they included patients qualitative synthesis (Figure 1). Duplicate authors were ≥41 years old (regardless if age groups were reported noted, but overlapping patient populations were not present separately) or if metastatic disease was present at upon detailed review. Included Eligibility Screening Identification Sarcoma 3 Table 1: Summary of the prognostic factors evaluated in the 9 studies included in the systematic review that included univariate or multivariate analysis. Bispo Junior ´ and Hauben Ferrari et al. Zhou et al. Ferrari et al. Kong et al. [14] Kim et al. [16] Ward et al. [19] Smeland et al. [12] Camargo [15] et al. [13] [17] [18] [20] LRR MFS MFS (N � 114) MFS (N � 24) MFS (N � 365) RFI (N � 19) MFS (N � 111) EFS (N � 1395) DFS (N � 127) (N � 33) (N � 19) UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA UVA MVA NS NS (15; Age NS (15) — NS (15) — S (2; 15; 40) S — — — — — — — S (sex dep) S — (q5y) 21; 30) Gender NS — NS — NS — NS — — — — — — — NS NS NS — Tumor location NS — NS — S S NS — — — — — NS — S S NS — Tumor size/ S S S (s � 6 cm; S (v � 1/3 S NS S — — — — — — — — S (v) NS — volume# (v � 150 ml) (s � 15 cm) 8 cm) bone) Osteosarcoma NS — S S NS — NS — — — — — NS — S NS NS — subtype Surgical margin S S NS — — — — — — — — — S S NS NS — — Histological S NS — — S S NS — — — — — S (90; 98) S S S S S response HER-2/neu — — — — — — — — — — S — — — — — — — expression P-glycoprotein — — — — — — — — NS — — — — — — — — ErbB-2 — — — — — — — — S — — — — — — — — — p53 — — — — — — — — S — — — — — — — — — BCL-2 — — — — — — — — — — — — — — — — — — Chemotherapy S S — — — — — — — — — — — — NS S — — protocol (MTX) (MTX) Lactate — — — — — — — — — — — — — — — — S S dehydrogenase Alkaline — — — — — — — — — — — — — — — — S NS phosphatase Pathologic NS — — — NS — — — — — — — — — NS NS fracture Limb salvage vs — — NS — — — — — — — — — — — — — amputation Time to seek — — NS — — — — — — — — — — — care UVA � univariate analysis; — � not tested; MFS � metastasis-free survival; MVA � multivariate analysis; � cut off points in years in parenthesis; q5y � 5-year increments; ns � not significant (P< 0.05); EFS � event-free survival; # � measure in parenthesis (s � significant; (s) � size and (v) � volume). 4 Sarcoma Tumor size/volume showed a slightly stronger associa- 3.1. Characteristics of Included Studies and Prognostic Factors. Most studies were retrospective in nature, and sample size tion than tumor location. *ree of the studies demonstrated an increased risk of metastases after MVA. However, there varied from 19 to 2680 patients. One study [12] was a prospective study. *is led to significant heterogeneity in were differences in both the type of measurements and sample sizes and also the majority of the sample sizes were cutoffs used for the analyses. Two of the studies used volume; small. All other articles’ subjects were derived from a single however, Kong et al. [14] used a hard cutoff of 150 mL while institutional patient population or from international col- Smeland et al. [12] evaluated the relative cutoff of 1/3 the laboration datasets [12, 13]. *ere was heterogeneity volume of the involved bone. Kim et al. [16] used maximal amongst the clinical prognostic factors that were evaluated tumor diameter in their evaluation and nomogram devel- for their association with subsequent metastasis, but several opment for the risk of metachronous metastases using factors were evaluated in the majority of the studies: age, evaluation of prognostic factors at diagnosis. *e histological subtype of these high-grade lesions did gender, tumor location, tumor size/volume, histologic subtype, and histological response (Table 1). Multivariate not appear to be significant in determining the risk of subsequent metastases. Seven of the studies performed a analysis (MVA) was performed in the reviewed papers when univariate significance was noted. UVA with only two showing an impact. After MVA, only Among these common factors, there was substantial Bispo Junior ´ and Camargo [15] showed osteoblastic subtype variation in definitions and cut off values used for the to have an effect on metastasis-free survival (MFS) with measurement of variables of interest. Age ranges varied worse rates of metastatic development compared to subtypes greatly among studies. Tumor location definitions were often other than osteoblastic. *is study did have a very low very broad or very specific; for example, Kong et al. [14] used sample of 24 patients and only stratified the subtypes as osteoblastic and nonosteoblastic. humeral vs other, while Ward et al. [19] delineated areas by bone and proximal vs distal. Studies were varied on the use of either tumor size or tumor volume measurements, and the 3.4. Surgical Margin/Histological Response. An uninvolved cutoffs used to assess risk varied study to study. Histology was surgical margin on pathological review has historically been evaluated as per pathological diagnosis in all instances, but considered important to improving outcomes in osteosar- like tumor location, the types were either somewhat broad coma [13, 16, 22]. Four studies evaluated the association with (osteoblastic vs nonosteoblastic) or somewhat specific (os- Kong et al. [14] and Ward et al. [19] showing significance teoblastic vs chondroblastic vs telangiectatic etc.). Percent with both UVA and MVA. Both of these studies used a necrosis was relatively consistent using a single stratification binary approach of involved/not-involved margins, while point of 90%; however, some studies further stratified above Bispo Junior ´ and Camargo [15] evaluated margins of 2 mm 90% while others further substratified based on the Salzer- and Smeland et al. [12] used qualitative measures of “wide/ Kuntschick grading scale [21]. radical, marginal, and intralesional.” *e studies that showed significance agree with previous cooperative group 3.2. Age/Gender. Most of the studies evaluated age at di- studies that have shown that the width of the margin does agnosis as a risk factor for the development of metachronous not appear to be significant as long as the definitive biopsy/ metastases. *ese studies used a variety of age cutoffs for resection are performed at a center with orthopedic on- cology experience [23, 24]. their analysis. Kong et al. [14] and Bispo Junior ´ and Camargo [15] used a single cutoff of 15 years, while Smeland Six of the studies evaluated the prognostic factor of histological response to neoadjuvant chemotherapy, and all et al. [12] had a gender specific age cutoff of either 11 or 12 years to distinguish between children and adolescents in used a necrosis/cell death percentage of at least 90% as the cutoff for evaluation [12–14, 16, 19, 20]. Five of the studies girls and boys, respectively. Hauben et al. [13] used 5-year increments to stratify the risk of metastatic recurrence. found continued significance on MVA. Kong et al. [14] Association of age with risk of metastases development was demonstrated a change on MFS on UVA while Hauben et al. not consistent among the studies; however, it was found to [13] did not. *is prognostic factor can be difficult to be a significant factor after MVA in EFS in the one pro- evaluate due to pathologist-to-pathologist variability in as- spective study that was included [12], demonstrating an sessments. Also, given the heterogeneity of the studies and improved survival for younger patients. Sex was included in neoadjuvant therapy regimens varied (even in a single in- stitutional study such as that by Ward et al. [19]) or were not 6 of the 9 studies, but did not show any prognostic sig- nificance with metastases development in univariate analysis clarified in the article [14]. Smeland et al. [12] performed a prospective study using the current standard of care of (UVA). methotrexate, doxorubicin, and cisplatin which is accepted as the standard of care after increasing the EFS and overall 3.3. Tumor Location/Tumor Size/Histological Subtype. survival to 54% and 71%, respectively. Location of the primary tumor was also evaluated in the majority of studies (7/9), but had variance in its estimated 3.5. Other Prognostic Factors. Several other factors were impact on metachronous metastases development. Studies by Kim et al. [16] and Smeland et al. [12] showed that evaluated among the studies but were not included in more primary tumor location in the proximal humerus had an than 3 of the 9 studies. Pathologic fracture (3 studies), limb increased risk of metastases. salvage vs amputation (1 study), time to seek care (1 study), Sarcoma 5 Primary tumor burden not only had a varying definition elevated alkaline phosphatase (ALP) (1 study), and increased Bcl-2 expression and increased P-glycoprotein expression (1 in the form of measurement criteria (volume vs size) but also had varying cutoffs between studies that did use similar study) did not show any significant changes in endpoint measurements in these individual studies. forms of measurement. As mentioned earlier, some used fixed measurements (150 ml) [14] while others used relative measurements (<1/3 of involved bone) [12]. *e latter would 4. Discussion be a more appropriate measure for pediatrics, given the *e goal of this review was to systematically evaluate the differences in potential size of the patients; however, it would literature from 1990 to present to delineate the clinical/ make data gathering potentially less consistent and more pathological risk factors that could be used to potentially subjective. predict which patients with localized osteosarcoma are more Interestingly, two molecular/genetic factors were found on UVA to have significance but were only evaluated in two likely to develop subsequent metastases. *e majority of studies were retrospective in nature, while some were studies in our search. *ere are several other molecular/ genetic factors that are found in the literature to potentially designed to address prognostic factors directly and others did not have that direct intent. *ese studies had cohorts that play a role in the outcome for patients with osteosarcoma, were able to be reviewed, but not included in a meta-analysis but consensus was clearly lacking. Based on our exclusion mainly due to the heterogeneity in definitions/cutoffs of the criteria, many of these studies were omitted as there was no different prognostic factors as well as treatment regimens data/analysis present on subsequent metastatic lesion de- being different or unclear. Also, while our initial intent was velopment or patients had metastases at diagnosis. In the to concentrate on pediatric patients (i.e.,<20 years of age at studies included in our review, Ferrari et al. [17] found that p53 mutations and absent ErbB-2 were associated with diagnosis), we subsequently opted to include studies that had patient populations less than 40 years of age due to rarity of decreased recurrence free interval (RFI), but only in UVA and in a small sample size of 19 patients. Zhou et al. [18] were the disease and to help increase the yield of the literature search. *e studies were limited to primary osteosarcoma in able to demonstrate that cytoplasmic HER-2 expression on osteosarcoma increased the risk of metastasis development. order to reduce the risk of confounding by secondary dis- ease. Histological response appeared to be the most con- HER-2 is currently being evaluated as a therapeutic target for sistent factor in prognosis for MFS in our study with osteosarcoma in a phase 1 study using chimeric antigen suboptimal response to neoadjuvant therapy (<90% ne- receptor T-cells [26]. While there may be therapeutic po- crosis), carrying a significant increase in the risk of me- tential with this receptor, its inclusion as prognostic factor tastases development. *e significance of this response is for the development of metachronous metastases cannot be echoed in the recent Children’s Oncology Group Study determined based on the current literature study. Of the factors that were found to not have significance in AOST0331 as the aim of the study was to intensify therapy in those patients with suboptimal response with the addition of the literature that was reviewed, presenting with a pathologic fracture at diagnosis was somewhat of a surprise. Pathologic ifosfamide and etoposide. However, therapy intensification did not improve outcomes and led only to increased toxicity fractures have been shown to potentially increase the risk of subsequent metastases while not affecting overall survival for the subjects [12]. Other factors that demonstrated significance via MVA [27]. However, many articles were not included in our study were age, tumor location, and tumor size/volume. Age is as the metastatic status at diagnosis was not specified or difficult to determine with any consistency among the there was no distinction between metachronous and syn- studies, given the different cutoffs that were used between chronous metastases. *is significantly limits our ability to studies. Also, our inclusion of those studies with older truly assess the potential risk that pathological fracture may patients could be skewing the data in those older age cutoff have on the development of metachronous metastases. *e development of a predictive nomogram by Kim et al. groups. Tumor location was not found to be a significant factor in [16] was an encouraging exercise and took prognostic factors into consideration that were consistent in significance with the majority of the studies it was evaluated in. Kim et al. [16] and Smeland et al. [12] showed significance for tumor lo- our review. While the nomogram appeared to be accurate in predicting actual EFS, it was only bootstrapped with cohorts cation; however, Smeland et al. [12] combined both proximal humerus and proximal femur making it more difficult to from the population that was used to develop the nomo- delineate the effect. Interestingly, these were the same two gram. Although the population was from a single institution, studies that demonstrated significance with regards to age, minimizing heterogeneity in several potential confounders raising the question of whether there is an association be- also led to a smaller sample size being used as the basis for tween age of presentation and site of primary lesion. *e the nomogram. Kim et al. [28] is another Korean group that variation of definitions of location used in the studies may also attempted to use a predictive nomogram and added in the use of ALP and capsular invasion. However, they had similar have made this risk factor more difficult to evaluate. Some studies had very broad definitions with respect to location, pitfalls with a small sample size of 141 patients and again consisted of a single ethnic group [28]. External validation of while others were more specific to sections of certain bones. *ese variations may explain why poor prognostic factors, these nomograms with larger and minimally heterogeneous such as axial skeletal and pelvic primary lesions, may have populations is required. *e establishment of cooperative been underappreciated in these studies [25]. study groups such as the European and American 6 Sarcoma other, telangiectatic, small cell, and high-grade Osteosarcoma Study (EURAMOS-1) will eventually lead to more standardized study populations and allow for these surface osteosarcoma [12]. prognostic factors to be evaluated in a prospective manner, (5) Surgical margin determination should be consistent potentially serving as the basis for new nomograms or as a using the Musculoskeletal Tumor Society (MSTS) confirmatory resource for established nomograms. originally defined by Enneking et al. [30], and this *e ability to predict the risk of metastatic disease in system classifies resection margins as intralesional patients with localized osteosarcoma would be beneficial for (macroscopic or microscopic tumor at the margin), several reasons. Most notably is a nomogram could be used marginal (resection through the pseudocapsule or for treatment stratification with the potential to spare low- reactive zone around the tumor), wide (the presence risk patients the considerable toxicity associated with our of normal tissue between tumor/pseudocapsule and current standard of care regimen. However, identification of margin), or radical (entire anatomic compartment these patients has been difficult, mainly due to issues with excised). identifying consistent prognostic factors to develop a (6) Histological response cutoff for adequacy of neo- stratification nomogram similar to the one developed by adjuvant chemotherapy should continue to be 90% Kim et al. [28] and Kim et al. [16]. During our literature tumor necrosis based on pathological determination. search and review, several factors became apparent that Central review of tissue samples should be incor- emphasize potential issues when performing a systematic porated in future clinical protocols to decrease as review or potential meta-analysis of a rare disease. First, the much interobserver variability from pathologist to number of patients between the studies varies greatly and pathologist as possible. often times can be insufficient to draw significant and consistent conclusions regarding a particular prognostic *e development of metastases is a difficult obstacle to factor. *is was true for the majority of the literature in- the successful treatment of osteosarcoma not only in per- cluded in our review, except for the Smeland et al. [12] study. sistence of disease but also in the need for more intensified Large collaborative studies such as EURAMOS-1 and those therapies. Being able to predict which patients have an conducted by the Children’s Oncology Group are absolutely increased risk of subsequent metastases after upfront necessary to obtain large pools of patients with rare diseases. therapy, would potentially allow for therapy stratification While it is important to perform these prospective studies, it after local control measures. Unfortunately, the literature is is critical to perform systematic reviews/meta-analyses to lacking in large-scale studies evaluating defined prognostic maximally use data from all available studies to draw more factors, limiting the creation of a predictive model of this accurate conclusions and inform subsequent trial rare disease difficult. *e movement towards large-scale development. collaborative studies between organizations and countries *e second major issue impacting the ability to combine will hopefully allow for the development of predictive analyses of published data is the discrepancy of measure- models in the future, ultimately leading to more effective ment cutoffs and the definitions used for several of the treatment. prognostic factors. We recommend that future studies use consistent definitions of the following prognostic factors Conflicts of Interest based on our literature review and experience: *e authors declare that they have no conflicts of interest. (1) Age should be used in sex-determined breaks due to typical differences in average ages of growth accel- References eration between the sexes. Smeland et al.’s [12] age cutoffs for children, adolescents, and adults were not [1] L. Mirabello, R. J. Troisi, and S. A. Savage, “Osteosarcoma too narrow as to cause smaller sample sizes but did incidence and survival rates from 1973 to 2004,” Cancer, distinguish differences in skeletal growth/maturity vol. 115, no. 7, pp. 1531–1543, 2009. levels. Also, Jun et al. [29] did demonstrate evidence [2] M. P. Link, A. M. Goorin, A. W. Miser et al., “*e effect of of increased growth velocity at diagnosis that was adjuvant chemotherapy on relapse-free survival in patients with osteosarcoma of the extremity,” New England Journal of associated with worse outcomes. Medicine, vol. 314, no. 25, pp. 1600–1606, 1986. (2) Tumor location should have categories of femur/ [3] E. E. Pakos, A. D. Nearchou, R. J. Grimer et al., “Prognostic tibia, other limb bones to include humerus, and axial factors and outcomes for osteosarcoma: an international skeleton. Location of proximal/distal and meta- collaboration,” European Journal of Cancer, vol. 45, no. 13, physis/diaphysis should also be included. pp. 2367–2375, 2009. [4] P. A. Meyers, C. L. Schwartz, M. Krailo et al., “Osteosarcoma: (3) Tumor size should be determined by a consistent a randomized, prospective trial of the addition of ifosfamide tumor length cutoff of <8 cm or ≥8 cm [16], with a and/or muramyl tripeptide to cisplatin, doxorubicin, and relative length of the involved bone (greater than/less high-dose methotrexate,” Journal of Clinical Oncology, vol. 23, than 1/3 of the bone) also included. no. 9, pp. 2004–2011, 2005. (4) *e osteosarcoma subtype should be categorized by [5] D. D. Rasalkar, W. C. W. Chu, V. Lee, B. K. Paunipagar, the histological subtypes of conventional chondro- F. W. T. Cheng, and C. K. Li, “Pulmonary metastases in blastic, conventional osteoblastic, conventional children with osteosarcoma: characteristics and impact on Sarcoma 7 patient survival,” Pediatric Radiology, vol. 41, no. 2, [20] S. Ferrari, G. Bacci, P. Picci et al., “Long-term follow-up and pp. 227–236, 2011 Feb. post-relapse survival in patients with non-metastatic osteo- [6] M. B. Harris, P. Gieser, A. M. Goorin et al., “Treatment of sarcoma of the extremity treated with neoadjuvant chemo- therapy,” Annals of Oncology, vol. 8, no. 8, pp. 765–771, 1997. metastatic osteosarcoma at diagnosis: a pediatric oncology group study,” Journal of Clinical Oncology, vol. 16, no. 11, [21] M. Salzer-Kuntschik, G. Delling, G. Beron, and R. Sigmund, pp. 3641–3648, 1998. “Morphological grades of regression in osteosarcoma after [7] G. Bacci, M. Rocca, M. Salone et al., “High grade osteosar- polychemotherapy? Study COSS 80,” Journal of Cancer Re- coma of the extremities with lung metastases at presentation: search and Clinical Oncology, vol. 106, no. S1, pp. 21–24, 1983. treatment with neoadjuvant chemotherapy and simultaneous [22] H. S. Kim, K. H. Shin, H. Y. Kim et al., “Postoperative no- resection of primary and metastatic lesions,” Journal of mogram to predict the probability of metastasis in Enneking Surgical Oncology, vol. 98, no. 6, pp. 415–420, 2008. stage IIB extremity osteosarcoma,” BMC Cancer, vol. 14, [8] K. A. Janeway and H. E. Grier, “Sequelae of osteosarcoma p. 666, 2014. medical therapy: a review of rare acute toxicities and late [23] A. L. Jun, M. S. Kim, D. H. Kim et al., “Osteosarcoma de- effects,” 1e Lancet Oncology, vol. 11, no. 7, pp. 670–678, 2010 veloped in the period of maximal growth rate have inferior Jul. prognosis,” Journal of Pediatric Hematology/Oncology, vol. 30, [9] A. M. Davis, R. S. Bell, and P. J. Goodwin, “Prognostic factors no. 6, pp. 419–424, 2008. [24] W. F. Enneking, S. S. Spanier, and M. A. Goodman, “A system in osteosarcoma: a critical review,” Journal of Clinical On- cology, vol. 12, no. 2, pp. 423–431, 1994. for the surgical staging of musculoskeletal sarcoma,” Clinical [10] G. Bacci, A. Longhi, M. Versari, M. Mercuri, A. Briccoli, and Orthopaedics and Related Research, vol. 153, pp. 106–120, P. Picci, “Prognostic factors for osteosarcoma of the extremity 1980. treated with neoadjuvant chemotherapy,” Cancer, vol. 106, [25] S. S. Bielack, B. Wulff, G. Delling et al., “Osteosarcoma of the no. 5, pp. 1154–1161, 2006. trunk treated by multimodal therapy: experience of the Co- [11] S. S. Bielack, B. Kempf-Bielack, G. Delling et al., “Prognostic operative Osteosarcoma study group (COSS),” Medical and factors in high-grade osteosarcoma of the extremities or Pediatric Oncology, vol. 24, no. 1, pp. 6–12, 1995. trunk: an analysis of 1,702 patients treated on neoadjuvant [26] National Library of Medicine (US), “Bethesda softworks,” cooperative osteosarcoma study group protocols,” Journal of National Library of Medicine (US), Bethesda, MD, Identifier: Clinical Oncology, vol. 20, no. 3, pp. 776–790, 2002. NCT00902044, Her2 Chimeric Antigen Receptor Expressing [12] S. Smeland, S. S. Bielack, J. Whelan et al., “Survival and T Cells in Advanced Sarcoma 2009 May 14 [cited 2019 May prognosis with osteosarcoma: outcomes in more than 2000 10]; [about 4 screens]. ClinicalTrials.gov [Internet], Available from: https://clinicaltrials.gov/ct2/show/NCT00902044, 2000. patients in the EURAMOS-1 (European and American Os- teosarcoma Study) cohort,” European Journal of Cancer, [27] Y. Zhou, Q. Lu, J. Xu et al., “*e effect of pathological fractures vol. 109, pp. 36–50, 2019. on the prognosis of patients with osteosarcoma: a meta- [13] E. I. Hauben, S. Bielack, R. Grimer et al., “Clinico-histologic analysis of 14 studies,” Oncotarget, vol. 8, no. 42, parameters of osteosarcoma patients with late relapse,” Eu- pp. 73037–73049, 2017. ropean Journal of Cancer, vol. 42, no. 4, pp. 460–466, 2006. [28] S. Jasnau, U. Meyer, J. Potratz et al., “Craniofacial osteosar- [14] C.-B. Kong, W. S. Song, W. H. Cho, J. M. Oh, and D.-G. Jeon, coma,” Oral Oncology, vol. 44, no. 3, pp. 286–294, 2008. “Local recurrence has only a small effect on survival in high- [29] D. Andreou, S. S. Bielack, D. Carrle et al., “*e influence of risk extremity osteosarcoma,” Clinical Orthopaedics and Re- tumor- and treatment-related factors on the development of lated Research, vol. 470, no. 5, pp. 1482–1490, 2012. local recurrence in osteosarcoma after adequate surgery. An [15] R. Z. Bispo Junior ´ and O. P. Camargo, “Prognostic factors in analysis of 1355 patients treated on neoadjuvant Cooperative Osteosarcoma Study Group protocols,” Annals of Oncology, the survival of patients diagnosed with primary non-meta- static osteosarcoma with a poor response to neoadjuvant vol. 22, no. 5, pp. 1228–1235, 2011. chemotherapy,” Clinics (Sao Paulo), vol. 64, no. 12, [30] N. M. Marina, S. Smeland, S. S. Bielack et al., “Comparison of pp. 1177–1186, 2009. MAPIE versus MAP in patients with a poor response to [16] M. S. Kim, S.-Y. Lee, T. R. Lee et al., “Prognostic nomogram preoperative chemotherapy for newly diagnosed high-grade for predicting the 5-year probability of developing metastasis osteosarcoma (EURAMOS-1): an open-label, international, after neo-adjuvant chemotherapy and definitive surgery for randomised controlled trial,” 1e Lancet Oncology, vol. 17, AJCC stage II extremity osteosarcoma,” Annals of Oncology, no. 10, pp. 1396–1408, 2016. vol. 20, no. 5, pp. 955–960, 2009. [17] S. Ferrari, F. Bertoni, L. Zanella et al., “Evaluation of P-gly- coprotein, HER-2/ErbB-2, p53, and Bcl-2 in primary tumor and metachronous lung metastases in patients with high- grade osteosarcoma,” Cancer, vol. 100, no. 9, pp. 1936–1942, [18] H. Zhou, R. L. Randall, A. R. Brothman, T. Maxwell, C. M. Coffin, and R. E. Goldsby, “HER-2/neu expression in osteosarcoma increases risk of lung metastasis and can Be associated with gene amplification,” Journal of Pediatric Hematology/Oncology, vol. 25, no. 1, pp. 27–32, 2003. [19] W. G. Ward, K. Mikaelian, F. Dorey et al., “Pulmonary metastases of stage IIB extremity osteosarcoma and subse- quent pulmonary metastases,” Journal of Clinical Oncology, vol. 12, no. 9, pp. 1849–1858, 1994.

Journal

Sarcoma – Hindawi Publishing Corporation

Published: Mar 27, 2020

There are no references for this article.